1. Field of the Invention
The present invention relates to recombinant, proteins composed of the assembly of the peptide sequences described for interacting with platelet collagen receptors. These proteins have a pro-aggregating activity, independent of the formation of a triple helix. They can be produced in bacteria and in mammalian cells.
2. Description of Related Art
Collagens are the principal structural components of the extracellular matrix of all multicellular organisms. They are a family of proteins composed of 28 different types that play a role during development and in tissue homeostasis. They are capable of being assembled into various supramolecular structures in the form of fibrils, microfibrils or networks.
Collagens have the common characteristic of containing one or more domains that have a triple helix structure formed of three polypeptide chains, or α chains, coiled together. As all three are amino acids, this characteristic is enabled by the presence of a glycine in the helicoidal motifs which consist of repeated G-X-Y sequences where X is often a proline and Y a hydroxyproline. This hydroxyproline is essential so the stabilisation of she triple helix and is characteristic of collagens. The praline residues are hydroxylated primarily by prolyl-4-hydroxylase (P-4-H) into 4-hydroxyproline. There is a second hydroxylase, prolyl-3-hydroxylase, which enables the hydroxylation of proline when it is in position X, whereas in position Y a praline hydroxylated by P-4-H is already found. In a collagen molecule, the alpha chains can be identical or different.
Collagen molecules are formed of helicoidal domains (or collagenous domains) flanked by non-helicoidal domains, called N- and C-propeptides. Recognition of the three a chains that form a molecule and the start of their assembly are under the control of the C-terminal end (C-propeptide). This process is carried out at the endoplasmic reticulum. Subsequently, the N- and C-propeptides are excised during collagen maturation, leaving short non-helicoidal sequences, telopeptides.
Collagens play an important role at the vessel walls by maintaining their integrity and their elasticity. This wall is formed of type-I collagen and type-III collagen, which are fibrillar, and of type-IV collagen, in network form. It should be noted that type-III collagen is also strongly expressed at atheromatous plaques. Furthermore, collagens are capable of modulating the functions of certain cells by direct interaction with specific cellular receptors. Thus, collagens, mainly types I and III, are powerful activators of platelet function.
Type-III collagen is a homotrimer composed of three al chains arranged in a triple helix. Two C-P-P triplets (potentially hydroxylated), present at the C-terminal end of the α1 chains, are then sufficient and necessary to the nucleation and the folding of the triple helix (Bulleid at al., EMBO J. 1997 Nov. 17; 16(22): 6694-701). On the other hand, these two triplets are not sufficient to enable the initial combining of the three chains. It is interesting to note that Bulleid et al. (1997) observed greater efficiency of triple helix formation when three triplets are maintained, and this efficiency is even greater than for the wild molecule. Several proteins are involved in the assembly process of monomeric chains: HSP47 (heat shock protein) and PDI (protein disulfide isomerase). C-propeptide, located at the C-terminal end of monomeric chains, appears involved in the alignment of a chains and the formation of the disulfide bridges necessary to the stabilisation of the protein (Bulleid et al., EMBO J. 1997 Nov. 17; 16(22): 6694-701). This C-propeptide is thus necessary only to ensure the combining of the monomeric chains. Various points should be remembered in relation to it                It consists of a discontinuous sequence of 15 amino acids which determines the specific standard assembly of α1 chains (Hulmes D J, J Struct Biol. 2002 January-February; 137(1-2): 2-70).        A coiled-coil domain located at the beginning of the C-propeptide is involved in the trimerisation of collagen III (Bulleid et al., EMBO J. 1997 Nov. 17; 16(22): 6694-701). This domain consists of four heptapeptides (McAlinden et al., J Biol Chem. 2003 Oct. 24; 278(43): 42200-7).        It contains eight cysteine residues which enable the formation of intra- and intercatenary disulfide bridges.        
The presence of disulfide bridges between chains at the C-telopeptide or C-propeptide is not necessary for the combining of the chains and the formation of the triple helix (Bulleid et al., Biochem J. 1996 Jul. 1; 317(Pt 1): 195-202). It should be noted that the experiments are undertaken with the N-propeptide. The notions of trimer and of alpha triple helix seem paradoxically independent. In the absence of N-propeptide, it seems judicious to leave either cysteine 2 of the C-propeptide, or the two cysteines of the C-telopeptide. This latter motif, called the knot sequence (GPCCG) (SEQ ID NO: 22), enables the formation of disulfide bridges only by virtue of a preliminary phenomenon of assembly and folding of the al chains (Boudko and Engel, J Mol Biol. 2004 Jan. 30; 335(5): 1289-97).
Whether of traumatic origin or the consequence of atherosclerosis, damage to the arterial wall is accompanied by the destruction of the vascular endothelium and the exposure of thrombogenic components such as collagen. This is followed by adhesion of platelets at the damaged site, in contact with surfaces rich in collagen or collagen fragments, their activation and the formation of a thrombus. The adhesion and stabilisation of the platelets in contact with this collagen are enabled by multiple interactions, of high affinities, between receptors present at the surface of the platelets and the collagen. This adhesion can occur indirectly following the binding of the A1 domain of the von Willebrand factor (vWF) with the platelet complex formed of the glycoproteins (Gp) GpIb-V-IX, itself bound to collagen by its A3 domain, or by direct interaction between platelet receptor and collagen. Several receptors can bind collagen molecules at highly specific peptide sequences, namely integrin α2β1, which plays an important role in the stabilisation of the platelet in contact with collagen, and GpVI, considered as the most important receptor for platelet activation. Another receptor, TIIICBP (type-III collagen-binding protein), has been described as capable of binding collagen directly.
Blood vessel flow rate is a major determinant that defines the type of platelet collagen receptors recruited. At an elevated flow rate, the platelets interact with collagen via vWF via the GpIb receptor and then they are activated by binding via GpVI; integrin (A21 only intervenes as a stabiliser of the platelet-collagen bond. At a low flow rate, the platelets bind to collagen via integrin α2β1, followed by binding to GpVI, thus leading to their activation, which is the step that prepares platelet aggregation. Similarly, other platelet collagen receptors are involved such as TIIICBP. In all cases, GpVI plays a mal or role in platelet activation.
Platelet adhesion is a process that is now dissociated from platelet activation. Indeed, many peptides, corresponding to short peptide motifs of collagen, are capable of inducing the adhesion of platelets without leading to their activation. Nevertheless, some of these peptides have the capacity to induce platelet activation and exhibit a so-called pro-aggregating activity. Their triple helix structure seems to be an essential prerequisite for this activity. It should be noted that when some of these peptides remain in monomeric form, they exhibit an anti-aggregating activity by a mechanism which remains to be identified but which could be the occupation of sites that become unavailable for native collagen.
Most of the work aimed at identifying the peptide sequences involved in the adhesion of platelets with type-III collagen use the fragment α1 (III) CB4 corresponding to the digestion fragment by CNBr exhibiting the highest aggregating activity.
Various peptide motifs have been described in the past few years as being capable of binding and activating platelets. They have been tested in the form of peptides of two possible natures:                similar to collagen, formed by the repetition of conserved GPO motifs repeated n times, not present in the native sequence of type-III collagen;        or corresponding to peptides formed of peptide motifs present in the α1(III)CB4 sequence, principally obtained by chemical synthesis (Farndale et al., Biochem Soc Trans. 2008 April; 36(Pt 2): 241-50).        
Integrin α2β1 is a receptor for collagens, laminin and other ligands in various cell types including endothelial cells and platelets. α2β1 binds collagen via its domain I at the peptide motifs present in the sequence of fibrillar collagens. Various motifs have been described with the highest affinity for GFOGER (SEQ ID NO: 11) present at the al chain of type-I collagen. For type-III collagen, and contrary to type-I collagen, it seems that several GXYGER (SEQ ID NO: 9) motifs are necessary for optimal binding, with GLOGER (SEQ ID NO: 12), GMOGER (SEQ ID NO: 13), GROGER (SEQ ID NO: 14) and GAOGER (SEQ ID NO: 15) as principal motifs (Kim et al., J Biol Chem. 2005 Sep. 16; 280(37): 32512-20). Other motifs such as GLOGEN (SEQ ID NO: 16) and GLKGEN (SEQ ID NO: 17) have been described but exhibit low affinity for α2β1 (Raynal et al., J Biol Chem. 2006 Feb. 17; 281(7): 3821-31). It should be noted that all of the peptides synthesised from these motifs exhibit an adhesion activity when they are organised in the form of a triple helix.
Glycoprotein VI (GPVI) is a 60-65 kDa type-I transmembrane glycoprotein belonging to the immunoglobulin (Ig) superfamily. It is constitutively expressed on the surface of platelets in the form of a non-covalent complex with the γ chain common to Ig receptors (FcRγ). The peptides described as interacting with this receptor are peptides similar to collagen formed from the repetition of 4 to 10 GPO triplets (Morton et al., Biochem J. 1995 Mar. 1; 306(Pt 2): 337-44 and Smethurst et al., J Biol Chem. 2007 Jan. 12; 282(2): 1296-304). Although comprising 10% GPO motifs, the maximum number of repetitions of this triplet does not exceed three in the native sequence of type-III collagen. It should be further noted here that the formation of the triple helix or of a polymeric form, obtained chemically following a modification of the cysteine or lysine residues, is essential to confer on these peptides a pro-aggregating activity. Moreover, the presence of a hydroxyproline in position 3 is essential to this activity. Conversely, when they are in monomeric form, these peptides exhibit an anti-aggregating activity (Asselin et al., Biochem J. 1999 Apr. 15; 339(Pt 2): 413-8). Recently, Jarvis et al. identified the peptide motif in the type-III collagen sequence that exhibits the highest adhesion activity. It is composed of residues GAOGLRGGAGPOGPEGGKGAAGPOGPO (SEQ ID NO: 18) located at amino acids 523 to 549 of al (III)CB4 (Jarvis et al., Blood. 2008 May 15; 111(10): 4986-96). This peptide motif is composed of three GPO, which are not consecutive and which seems to be the minimal number necessary for proper interaction with GpVI.
Indirect interaction between the platelet receptor GpIb and collagen is dependent on vWF, which is a plasma multimeric protein secreted by endothelial cells and platelets in response to vascular damage or following an increase in parietal pressure. It plays a major role in the recruitment of platelets at damaged sites of vascular territories subjected to elevated flow rates. This factor is composed of three domains named A1, A2 and A3. It binds collagen via its A3 domain and binds the GpIb receptor by its A1 domain. Type-III collagen seems to have a single site, of high affinity for the A3 domain of vWF, which is also present in type-II collagen. This peptide motif is located between amino acids 403 and 413 and is composed of GPRGQOVMGFO (SEQ ID NO: 19) with certain amino acids critical for binding vWF (Lisman T et al., Blood. 2006 Dec. 1; 108(12): 3753-6). Verkleij et al. identified as a potential binding site amino acids 541 to 558 composed of GAAGPOGPOGSAGTOGLQ (Verkleij et al., Blood. 1998 May 15; 91(10): 3808-16; SEQ ID NO:20). This peptide motif is located between the vWF motif described by Lisman et al. and the GMOGER (SEQ ID NO: 13) integrin α1β2 binding motif.
The Fauvel-Lafève team described the existence of an octapeptide, KOGEOGPK (SEQ ID NO: 21), located between amino acids 655 and 662 of the fragment al (III)CB4. The platelet receptor recognised by this octapeptide has been identified and named TIIICBP (type-III collagen binding protein) (Monnet E et al., J Biol Chem. 2000 Apr. 14; 275(15): 10912-7). This octapeptide is capable of inhibiting the interaction of platelets with type-III collagen, but not with type-I collagen, under both static and flow conditions. More recently, Pires et al. showed that this octapeptide has an inhibiting activity on aggregation only when it is in homotrimer form. On the contrary, its structure in triple helix form by adding at these ends cysteines and GPP gives it a pro-aggregating activity (Pires et al., Eur J Med Chem. 2007 May; 42(5): 694-701).
The biological and ultrastructural properties of collagens, and notably their capacity of binding to membrane receptors, open a broad field of applications because of their multiple roles in tissues. However, these applications only have meaning if it is possible to have homogeneous preparations of these collagens in large quantities and in a reproducible way.
Two methods of production were used to this end. The work and developments conducted in these two areas are generally highly targeted for a given application. Thus, chemical synthesis can produce short peptides that represent only a tiny part of the protein, in general peptide motifs of interest. Its principal application relates to haemostasis and more precisely to modulation of the bond between collagen and platelets.
However, these peptides have in general only one of the activities sought and this can depend on the three-dimensional conformation of the peptide and notably of she formation of a triple helix. Thus, there is a significant need for the development of functionalised recombinant collagen proteins that can be produced by biological systems offering high productivity. A major obstacle is the difficulties of production and purification of these collagen-derived proteins because of their tendency to aggregate and to adhere.
To date, short peptide sequences (fewer than 50 amino acids) have all been produced by chemical synthesis and not via cellular production systems (Farndale et al., Biochem Trans. 2008 April; 36(Pt 2): 241-50). Moreover, these motifs have been synthesised in an isolated way. Conversely, the synthesis of recombinant type-III collagen, exhibiting platelet binding activity, has been carried out via cellular production systems and it is the totality of the sequence, for example proα1(III), which has been used. The objective sought is the synthesis of a whole procollagen capable of being organised into a triple helix (WO 9 307 889). These recombinant collagens have in principle multiple applications, namely those of the complete collagen sequence.
It is thus a question of synthesising smaller proteins of simpler structure that are thus less demanding in terms of production but that preserve the biological activities of interest of the native protein.